All Optical Super Resolution and Uncertainty Visualization

Official pytorch implementation of the papers:

• "Can the Success of Digital Super-Resolution Networks Be Transferred to Passive All-Optical Systems?" [Paper]

• "All-optical uncertainty visualization for ill-posed image restoration tasks" [Paper]

System Requirements

Hardware Requirements

This code has been tested on a Linux machine (Ubuntu 22.04.4 LTS) with NVIDIA GeForce GTX 2080 Ti GPU.

Software Requirements

This code has the following dependencies:

  python >= 3.8.12 
  torch >= 1.12.1
  torchvision >= 0.13.1
  numpy >= 1.23.4
  tqdm >= 4.64.0
  torchmetrics >= 1.4.0

Setting an environment

Create a python virtual environment, install all dependecies using the requirements.txt file and then run the code on your computer.

cd DIR_NAME
python3 -m venv VENV_NAME
source VENV_NAME/bin/activate
pip install -r requirements.txt 

Installation time should take around 10 minutes.

Usage Instructions

After installation one can run our code.

Data

The data used in our work is the MNIST, FashionMNIST, Quick, Drae!, KMNIST and EMNIST datasets.

The MNIST, FashionMNIST and KMNIST datasets are available via torchvision. Quick, Draw! is avilable via their official github. EMNIST dataset is available via their official website.

For both the KMNIST and EMNIST datasets we used the drop-in replacemnt for the MNIST dataset.

Hyperparameters

config.py include all the hyperparameters used for each trial. The different hyperparameters used for running different experiemnts are detailed in the paper.

Usage for All-Optical Super-Resolution

python3 main_trials.py --epochs 1000 --lr 1e-1 --trial_name unique_sr_trial_name

Usage for Uncertainty Visualization

python3 Uncertainty-Visualization/main_trials.py --epochs 1000 --epcoh_initialize 500 --lr 1e-1 --trial_name unique_uncertainty_visualization_trial_name

For additional details see the README file in the sub direcorty.

Licence

Our code is under the MIT License.

Citation

If you use this code for your research, please cite our paper:

@article{kleiner2025can,
  title={Can the success of digital super-resolution networks be transferred to passive all-optical systems?},
  author={Kleiner, Matan and Michaeli, Lior and Michaeli, Tomer},
  journal={Nanophotonics},
  volume={14},
  number={19},
  pages={3181--3190},
  year={2025},
  publisher={De Gruyter}
}

@article{kleiner2026all,
  author = {Kleiner, Matan and Michaeli, Tomer},
  journal = {Opt. Lett.},
  number = {8},
  pages = {2012--2015},
  publisher = {Optica Publishing Group},
  title = {All-optical uncertainty visualization for ill-posed image restoration tasks},
  volume = {51},
  month = {Apr},
  year = {2026},
}